8. OpenAI Responses Target

8. OpenAI Responses Target#

In this demo, we show an example of the OpenAIResponseTarget. Responses is a newer protocol than chat completions and provides additional functionality with a somewhat modified API. The allowed input types include text, image, web search, file search, functions, reasoning, and computer use.

Before you begin, ensure you are set up with the correct version of PyRIT installed and have secrets configured as described here.

OpenAI Configuration#

Like most targets, all OpenAITargets need an endpoint and often also needs a model and a key. These can be passed into the constructor or configured with environment variables (or in .env).

endpoint: The API endpoint (OPENAI_RESPONSES_ENDPOINT environment variable). For OpenAI, these are just “https://api.openai.com/v1/responses”.
auth: The API key for authentication (OPENAI_RESPONSES_KEY environment variable).
model_name: The model to use (OPENAI_RESPONSES_MODEL environment variable). For OpenAI, these are any available model name and are listed here: “https://platform.openai.com/docs/models”.

from pyrit.common import IN_MEMORY, initialize_pyrit
from pyrit.executor.attack import ConsoleAttackResultPrinter, PromptSendingAttack
from pyrit.prompt_target import OpenAIResponseTarget

initialize_pyrit(memory_db_type=IN_MEMORY)

target = OpenAIResponseTarget()

attack = PromptSendingAttack(objective_target=target)

result = await attack.execute_async(objective="Tell me a joke")  # type: ignore
await ConsoleAttackResultPrinter().print_conversation_async(result=result)  # type: ignore

────────────────────────────────────────────────────────────────────────────────────────────────────
🔹 Turn 1 - USER
────────────────────────────────────────────────────────────────────────────────────────────────────
  Tell me a joke

────────────────────────────────────────────────────────────────────────────────────────────────────
🔸 ASSISTANT
────────────────────────────────────────────────────────────────────────────────────────────────────
  Why did the scarecrow win an award?
    Because he was outstanding in his field!

────────────────────────────────────────────────────────────────────────────────────────────────────

Tool Use with Custom Functions#

In this example, we demonstrate how the OpenAI Responses API can be used to invoke a custom-defined Python function during a conversation. This is part of OpenAI’s support for “function calling”, where the model decides to call a registered function, and the application executes it and passes the result back into the conversation loop.

We define a simple tool called get_current_weather, which simulates weather information retrieval. A corresponding OpenAI tool schema describes the function name, parameters, and expected input format.

The function is registered in the custom_functions argument of OpenAIResponseTarget. The extra_body_parameters include:

tools: the full OpenAI tool schema for get_current_weather.
tool_choice: "auto": instructs the model to decide when to call the function.

The user prompt explicitly asks the model to use the get_current_weather function. Once the model responds with a function_call, PyRIT executes the function, wraps the output, and the conversation continues until a final answer is produced.

This showcases how agentic function execution works with PyRIT + OpenAI Responses API.

from pyrit.common import IN_MEMORY, initialize_pyrit
from pyrit.models import PromptRequestPiece, PromptRequestResponse
from pyrit.prompt_target.openai.openai_response_target import OpenAIResponseTarget

initialize_pyrit(memory_db_type=IN_MEMORY)


async def get_current_weather(args):
    return {
        "weather": "Sunny",
        "temp_c": 22,
        "location": args["location"],
        "unit": args["unit"],
    }


# Responses API function tool schema (flat, no nested "function" key)
function_tool = {
    "type": "function",
    "name": "get_current_weather",
    "description": "Get the current weather in a given location",
    "parameters": {
        "type": "object",
        "properties": {
            "location": {"type": "string", "description": "City and state"},
            "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]},
        },
        "required": ["location", "unit"],
        "additionalProperties": False,
    },
    "strict": True,
}

# Let the model auto-select tools
target = OpenAIResponseTarget(
    model_name="o4-mini",
    custom_functions={"get_current_weather": get_current_weather},
    extra_body_parameters={
        "tools": [function_tool],
        "tool_choice": "auto",
    },
    httpx_client_kwargs={"timeout": 60.0},
    # api_version=None,  # this can be uncommented if using api.openai.com
)

# Build the user prompt
prompt_piece = PromptRequestPiece(
    role="user",
    original_value="What is the weather in Boston in celsius? Use the get_current_weather function.",
    original_value_data_type="text",
)
prompt_request = PromptRequestResponse(request_pieces=[prompt_piece])

response = await target.send_prompt_async(prompt_request=prompt_request) # type: ignore

for idx, piece in enumerate(response.request_pieces):
    print(f"{idx} | {piece.role}: {piece.original_value}")

0 | assistant: The current weather in Boston is Sunny with a temperature of 22°C.

Using the Built-in Web Search Tool#

In this example, we use a built-in PyRIT helper function web_search_tool() to register a web search tool with OpenAI’s Responses API. This allows the model to issue web search queries during a conversation to supplement its responses with fresh information.

The tool is added to the extra_body_parameters passed into the OpenAIResponseTarget. As before, tool_choice="auto" enables the model to decide when to invoke the tool.

The user prompt asks for a recent positive news story — an open-ended question that may prompt the model to issue a web search tool call. PyRIT will automatically execute the tool and return the output to the model as part of the response.

This example demonstrates how retrieval-augmented generation (RAG) can be enabled in PyRIT through OpenAI’s Responses API and integrated tool schema.

NOTE that web search is NOT supported through an Azure OpenAI endpoint, only through the OpenAI platform endpoint (i.e. api.openai.com)

from pyrit.common import IN_MEMORY, initialize_pyrit
from pyrit.common.tool_configs import web_search_tool
from pyrit.models import PromptRequestPiece, PromptRequestResponse
from pyrit.prompt_target.openai.openai_response_target import OpenAIResponseTarget

initialize_pyrit(memory_db_type=IN_MEMORY)

target = OpenAIResponseTarget(
    extra_body_parameters={
        "tools": [web_search_tool()],
        "tool_choice": "auto",
    },
    httpx_client_kwargs={"timeout": 60},
)

prompt_piece = PromptRequestPiece(
    role="user", original_value="Briefly, what is one positive news story from today?", original_value_data_type="text"
)
prompt_request = PromptRequestResponse(request_pieces=[prompt_piece])

response = await target.send_prompt_async(prompt_request=prompt_request) # type: ignore

for idx, piece in enumerate(response.request_pieces):
    # Reasoning traces are necessary to be sent back to the endpoint for function calling even if they're empty.
    # They are excluded here for a cleaner output. 
    if piece.original_value_data_type != "reasoning":
        print(f"{idx} | {piece.role}: {piece.original_value}")

| assistant: {"id":"ws_68b78ebf92b481a299ee79eda74c5ae80cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"search","query":"positive news September 3 2025"}}
| assistant: {"id":"ws_68b78ec2147081a2b2078a9bcda00f9e0cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"open_page","url":"https://newsroom.ocfl.net/media-advisories/press-releases/2025/09/a-spontaneous-sale-for-a-happy-tail-orange-county-animal-services-reduced-adoption-fees-for-three-days/"}}
| assistant: {"id":"ws_68b78ec4823881a2a18c0ed68897d5fd0cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"search","query":"site:apnews.com good news September 3 2025"}}
| assistant: {"id":"ws_68b78ec7107081a28908f0f4997dfc5c0cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"search","query":"AP News One Good Thing September 3 2025"}}
| assistant: {"id":"ws_68b78ec99cb881a28420fdd84cf98b540cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"open_page","url":"https://apnews.com/article/biden-charleston-west-virginia-715c1faa116d4285c77f0095d37a714a"}}
| assistant: {"id":"ws_68b78ecb17a481a2a08b8d38954539580cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"find_in_page","pattern":"South Charleston"}}
| assistant: {"id":"ws_68b78ece497081a2a6633cdd4403ddcb0cc9740a19fd8e2f","type":"web_search_call","status":"completed","action":{"type":"search","query":"West Virginia electric school buses plant September Charleston GreenPower Motor Co. AP News"}}
| assistant: A Canadian electric-school-bus manufacturer, GreenPower Motor Co., will begin producing zero-emission buses at its new South Charleston, West Virginia, facility in September, Gov. Jim Justice announced today. The 9.5-acre plant will initially create about 200 jobs, with the capacity to expand to roughly 900 positions once full production is reached in two years ([apnews.com](https://apnews.com/article/biden-charleston-west-virginia-715c1faa116d4285c77f0095d37a714a?utm_source=openai)).

This expansion is supported by President Biden’s $1 trillion infrastructure law, under which the Environmental Protection Agency’s Clean School Bus Program offers rebates of up to $375,000 for larger electric buses and $285,000 for smaller models, plus an extra $20,000 per bus to fund charging infrastructure. GreenPower plans to leverage these rebates to help school districts transition away from diesel-powered fleets, with buses capable of 140 miles on a single charge and a recharge time as short as 3.5 hours ([apnews.com](https://apnews.com/article/biden-charleston-west-virginia-715c1faa116d4285c77f0095d37a714a?utm_source=openai)).

This development represents a win on multiple fronts: it accelerates the shift toward cleaner school transportation, reduces local air pollution, and brings significant employment opportunities to the Charleston area—all announced on September 3, 2025.

8. OpenAI Responses Target

Contents

8. OpenAI Responses Target#

OpenAI Configuration#

Tool Use with Custom Functions#

Using the Built-in Web Search Tool#